Tricky analyte, challenging matrix and a new high sensitivity analytical platform

Page created by Michael Goodwin
 
CONTINUE READING
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Tricky analyte, challenging matrix
and a new high sensitivity analytical platform:

                                          How to overcome major challenges for a successful
                                       biomarker assay validation on the SMCxPRO® platform

Katharina Schutz, Alessandra Bühler, Eginhard Schick, Stéphanie Vauléon
Roche Pharma Research and Early Development, Bioanalytical R&D, Roche Innovation Center Basel
13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Challenging the Quantification Limit of LBAs: Case Study
Introduction

Tricky analyte, challenging matrix
• Biomarker for neurodegenerative disease in human cerebrospinal fluid (CSF)
• Analyte heterogenous in length among patients & prone to aggregation
• Low pg/mL concentration (low fM range)
• Exploratory assay available on the Erenna® platform (Merck) in a research laboratory

Context of use: core surrogate biomarker in late stage clinical development:
• Assay transfer to a regulated laboratory mandatory
• Full assay validation required
• Highest possible sensitivity to be achieved
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Assay Transfer Strategy

                                                              Q1 - Q2 2019

       Research lab (CRO)                                        failed               Regulated lab (CRO)
       Assay qualification                                                             Full assay validation
                                                                                     Clinical sample analysis

                                             Q3 2019 -
                                             Q2 2020
            Erenna®:
          end of support                                                                       SMCxPro®
        announced for 2021                                                                introduced in 2018

                                                         Roche internal Reg BA lab
                                                         Establishment of platform
                                                            Full assay validation
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Single Molecule Counting Technology
Erenna® Platform

• Ultra high sensitivity platform
  originally developed by Singulex®
• Manual bead-based sandwich
  immunoassay in 96-well plate
  format
• Elution of detection antibodies
  from immune complexes
• Quantification via capillary flow
  fluorescence detection

S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Single Molecule Counting Technology
SMCxPro® Platform

• Same bead-based immunoassay
  as Erenna® (identical kits)
• Readout in 384-well plate
  with a rotating laser allowing
  for individual photon counting
• Sophisticated laser optic: daily
  self-calibration and weekly
  external calibration of instrument

S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Robustness of the SMCxPRO® Platform
Device Installation Jun – Oct 2019

Technical challenges
• No proper assay signals obtained during on site analyst training using IL6 assay kit
• Red flags popped up at calibration: several visits of engineer required for laser & software adjustment
• The fixed, automatic distinction between background noise and event (event threshold) lead to artifacts
  at very low signal levels. Demo software update offered.

à Intense information exchange with Merck team to solve issues
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Robustness of the SMCxPRO® Platform
Device Performance Monitoring

Systematic checks implemented to discriminate between assay problems and device related issues
• Reading plate (96-well quadrant) loaded with unique concentration of detection antibody
  providing a low assay signal and stored refrigerated for 1 month
• Daily analysis and signal precision across plate calculated
• Plate precision ranged between 6-11%; higher compared to ELISA

Mitigation: Samples analyzed in triplicates, possibility to remove one outlier, replicate precision to be ≤20%
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Robustness of the SMCxPRO® Platform
Pipetting & Bead Handling

Non-automated high sensitivity assay
•    Each assay step crucial
•    Visual check of proper bead peletting
     & resuspension is essential
•    Pipetting out of biosafety cabinet to avoid
     air flows (except biosample preparation)

         Proper                         Pipetting of bead solution on        Homogenization of bead         Accuracy of
    resuspension of                       96 well plate with manual         solutions during incubation        plate
       bead stock                        pipette instead of electronic   steps: homogenous shaking over     transfer (10
       (volume of                         pipettes (multi-stepper) to         the plate to be ensured       μL out of 12
        aliquots)                         avoid plate inhomogeneity         (selection of shaker crucial)       μL!)
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Robustness of the SMCxPRO® Platform
Microplate Washer Technology

• SMCxPro delivered with BioTek 405 TS Plate Washer:
  aspiration washer used with a magnetic carrier plate

• Washer settings pre-adjusted at manufacturing
  but not re-adjusted/checked at installation (Merck
  engineers now trained)

• Device adjusted by BioTek in Nov 2019: performance
  improved but still inhomogeneity throughout plate

                                                         Assay signal,
                                                         unique concentration
                                                         of analyte loaded
                                                         all over the plate
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Tricky analyte, challenging matrix and a new high sensitivity analytical platform
Robustness of the SMCxPRO® platform
Microplate Washer Technology

• Centrifugal Blue®Washer designed
  for cell- and bead-based assays

• Wash step example:
  – Plate 2 min on magnetic carrier
  – Buffer spinned out at 800 rpm
  – 2 cycles of buffer dispense
    & centrifugation

• Optimization of wash programm
  required

• Better signal homogeneity obtained:
  Blue®Washer used from there

S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Robustness of the SMCxPRO® Platform
Accuracy and Precision Data After Optimization – Research Grade Reagents

Calibration samples
                       Nom                                                                        S/N at   S/N at
   Run Date                         0.00   1.63    4.08   10.2    25.6    64.0    160     400
                   Conc [pg/mL]                                                                   LLOQ     ULOQ
  19-Dec-2019       Signal [RU]     1.60    5.47   10.6   23.1    56.2     114     260     407      3.4     254

   07-Jan-20
                   Conc [pg/mL]
                    Signal [RU]     1.84
                                            1.63
                                            5.59
                                                   4.09
                                                   10.5
                                                          10.1
                                                          27.6
                                                                  26.7
                                                                  69.1
                                                                          59.6
                                                                           165
                                                                                   174
                                                                                   437
                                                                                           386
                                                                                           828     3.0      450
                                                                                                                    QC samples
                   Conc [pg/mL]             1.67   3.88   10.70   26.0    60.4     169     394                                                 Back-calculated concentrations at QC level [pg/mL]
   07-Jan-20        Signal [RU]     NV      7.80   15.7   35.2    91.6    186.0   550.0   1477     NA       NA             Run Date                 LQC              MQC                HQC
                   Conc [pg/mL]             1.59   4.18   10.4    27.8    56.2     162     423                                                     4.50              40.00               300
   09-Jan-20        Signal [RU]     2.83    6.39   12.7   29.3    72.1     169     409    1184     2.3      418           19-Dec-19                4.56              38.8                290
                   Conc [pg/mL]             1.62    4.1   10.4    26.5      62     150     428                                                     4.16              45.0                258
   10-Jan-20        Signal [RU]     2.02    4.37    7.7   16.7    37.7    85.9     230     601     2.2      298           07-Jan-20                5.16              46.5                308
                   Conc [pg/mL]             1.67    4.0   10.3    25.3    59.7     163     431                                                     4.95              46.1                304
   17-Jan-20        Signal [RU]     2.02    8.12   16.6   35.5    82.3     188     566    1272     4.0      630            7-Jan-20                4.74              40.3                312
                   Conc [pg/mL]             1.57    4.3   10.4    25.2    58.6     176     397                                                     6.21              47.2                301
   20-Jan-20        Signal [RU]     3.09    8.22   16.9   32.6    90.6     220     488    1343     2.7      435            9-Jan-20                6.37              32.1                321
                   Conc [pg/mL]             1.61    4.4    9.3    27.3    67.4     150     411                                                     4.39              40.8                299
   21-Jan-20        Signal [RU]     3.54    11.1   18.8   45.7    102.0    300     749    1730     3.1      489            10-Jan-20               5.06              40.0                295
                   Conc [pg/mL]             1.64    4.0   10.9    24.1    66.1     160     399                                                     4.88              44.1                336
   21-Jan-20        Signal [RU]     6.19    11.5   22.3   67.2    164.0    388     954    2320     1.9      375            20-Jan-20               5.69              43.5                364
                   Conc [pg/mL]             1.72   3.60   11.2    27.1    63.5     156     396                                                     5.28              48.4                337
             Mean Conc [pg/mL]              1.64   4.05   10.4    26.2    61.5     162     407                             21-Jan-20               4.65              40.0                352
         Interbatch Accuracy [%]           100.4   99.3   102.1   102.4   96.1    101.4   101.8                                                    4.64              45.5                325
        Interbatch Precision [%]            2.8     5.8    5.2     4.5     5.9     5.8     4.0                             21-Jan-20               4.09              37.1                297
                  Total error [%]           3.2     6.5    7.2     6.9     9.8     7.2     5.9                                                     4.86              37.5                284
                                                                                                                        Mean Conc [pg/mL]          4.98              42.1                311
                                                                                                                    Interbatch Accuracy [%]        110.7             105.1              103.8
QC sample data acknowlege for signal homogeneity over the plate                                                     Interbatch Precision [%]
                                                                                                                             Total error [%]
                                                                                                                                                   13.1
                                                                                                                                                   23.8
                                                                                                                                                                     10.6
                                                                                                                                                                     15.7
                                                                                                                                                                                         8.6
                                                                                                                                                                                        12.4

S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Towards Assay Validation
Labeling of Critical Reagents

Initial procedure: use of SMC labeling kits
• Black box approach, no quality control for product available
• Batch-to-batch variability observed

  Optimization at Roche Diagnostics
• Analytical characterization before and after labeling          Performance of labeled antibodies
  (purity, incorporation rate, fluorescence emission,            •   S/N ratio at LLOQ
  functional testing)                                                o   Labeling with kits: 2 to 4
• Buffer exchange/desalting optimized                                o   Optimized reagents: 5 to 7
• Variation of labeling ratio to enhance S/N ratio in assay      •   Long term stability assessed
                                                                 •   CoA generated
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Towards Assay Validation
Assay Matrix

 Sample collection
 • Sample analysis in triplicates, 135 μL sample per well: 500 μL aliquots required
 • Samples shock frozen directly after preparation to ensure analyte integrity

 Surrogate assay matrix
  • Large volume of rare matrix: surrogate assay matrix for preparation of calibration and QC samples
  • Ready to use artificial CSF

  Human CSF
  • Commercially available human CSF issued from leftover samples: uncontrolled chain of custody
  • Samples gave high background à selectivity demonstrated via parallelism assessement
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Towards Assay Validation
Reference Standard

 • Analyte heterogenous in length among
   patients: surrogate reference standard of
   defined length

 • Suitability of surrogate reference standard
   demonstrated on
         – recombinant fragments: response of
           fragments of different lengths parallel to
           reference standard curve
         à relative quantitative assay

                                                                 Red: Surrogate reference standard (different lots)
         – patient samples via parallelism                     Black: Representative recombinant forms of analyte
           experiment                                   Grey: Wild type fragments (not detected at relevant concentrations)

S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Towards Assay Validation
Validation Strategy

  Mitigation of assay variability
  • Samples analyzed in triplicates, possibility to remove one outlier, precision to be ≤20%
  • Acceptance criteria extended from ±20%/±25% A&P to ±30% based on pre-validation data

 Validation parameters
      – Inter- and intra-assay accuracy and precision     – Plate homogeneity
        on QC samples (surrogate matrix)                  – Hook effect
      – Inter-assay precision on patient samples          – Interferences
      – Parallelism on patient samples                    – Stability in surrogate matrix
      – Determination of LOD                              – Incurred sample stability

 Validation results
 • In house validation completed within one month. All pre-set criteria met. Target LLOQ validated
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Challenging the Quantification Limit of LBAs: Case Study Conclusion

• Discrimination between technical and analytical challenges
  allowed for successful assay optimization & validation               GxP lab (CRO)
                                                                    Full assay validation
• SMCxPro platform requires lab excellence:
       – extended control of devices
       – experienced & trained analysts

• Communication between manufacturers and labs key factor for
  implementation of new analytical platforms

• Deep in-house assay understanding allowed for efficient trouble
  shooting at CRO: external assay validation successfully
  completed                                                                 Q1-Q3
                                                                             2020

S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Acknowledgement

    pRED Bioanalytical R&D                         Roche Diagnostics
    • Katharina Schutz                             • Bernhard Maximilian Roettig, Tobias Oelschlaegel
    • Alessandra Bühler                            pRED High Throughput Screening
    • Eginhard Schick                              • Philippe Hartz
    • Benoit Massonnet                             Clinical Development Project Team
    • Julian Meier
    • Martin Schaefer                              Merck
    • Nicole Justies                               • Olivier Weyel, Robert Hardcastle, Christian Haag
    • Jasna Canadi                                 BlueCatBio
    • Matthew Barfield                             • Wolfgang Mann
    • Julia Heinrich                               CROs Lab Teams
S. Vauléon, 13th EBF Open Symposium, 17-Nov-2020
Doing now what patients need
           next
You can also read